Sex and Racial and Ethnic Diversity Among Ophthalmology Subspecialty Fellowship Applicants

IMPORTANCE: Physician-patient concordance in sex and race is associated with improved patient outcomes. Studies have explored diversity among ophthalmology residents and faculty, but to our knowledge, not among ophthalmology fellows. OBJECTIVE: To assess diversity by sex and race and ethnicity among fellowship applicants in ophthalmology subspecialties and compare match rates by applicants’ sex and underrepresented in medicine (URiM) status. DESIGN, SETTING, AND PARTICIPANTS: This cohort study examined ophthalmology subspecialty fellowship data from the 2021 San Francisco Match. MAIN OUTCOMES AND MEASURES: Applicant characteristics were stratified by sex and URiM status and compared using χ2, Mann-Whitney U, and median tests. For applicants who matched, the percentages of female and URiM applicants were compared among the ophthalmic subspecialties. A multivariable logistic regression model was used to assess the association of applicant characteristics with their match outcomes. RESULTS: Included in the sample were 537 candidates who applied for an ophthalmology fellowship using the 2021 San Francisco Match; 224 applicants (42.6%) were female, and 60 applicants (12.9%) had URiM status. Females and males had similar match rates (70.5% [n = 158] and 69.2% [n = 209], respectively; P = .74), but females had a higher median (IQR) US Medical Licensing Examination (USMLE) Step 2 Clinical Knowledge (CK) score (248 [240-258] vs 245 [234-254]; P = .01). The pediatric ophthalmology subspecialty had the highest percentage of female matched applicants (67.5%; 27 of 40 matched applicants), while the retina subspecialty had the highest percentage of males (68.9%; 84 of 122 matched applicants). URiM applicants had lower match rates (55.0%, n = 33) than non-URiM applicants (72.2%, n = 293; P = .007). The URiM applicants had lower median (IQR) scores on the USMLE Step 1 (238 [227-247]) compared with Asian applicants (246 [235-254]) and White applicants (243 [231-252]; P = .04). Additionally, URIM applicants submitted fewer median (IQR) applications (10 [1-23]) than Asian (21 [8-37]) and White (17 [8-32]; P = .001) applicants and completed fewer interviews (median [IQR], 2 [0-11]) than Asian (median [IQR], 12 [3-18]) and White applicants (median [IQR], 8 [1-14]; P = .001). Among matched fellows in each subspecialty, URiM applicants comprised 13.9% (n = 11) in glaucoma, 10% (n = 4) in pediatric ophthalmology, 7.3% (n = 6) in cornea, and 6.6% (n = 8) in retina. CONCLUSIONS AND